.sc iencedi rect .com

HO ST E D BY Avai lab le at www

INFORMATION PROCESSING IN AGRICULTURE 2 (2015) 64–71

journa l homepage: www.elsevier .com/ locate / inpa

Drying kinetics of technical specified rubber

http://dx.doi.org/10.1016/j.inpa.2015.05.0012214-3173 � 2015 China Agricultural University. Production and hosting by Elsevier B.V. All rights reserved.

* Corresponding author. Tel.: +60 017 4598849, +60 (3) 8924 8169(office); fax: +60 3 89248017.

E-mail address: kebx4nmn@nottingham.edu.my (L.C. Lim).

Peer review under responsibility of China Agricultural University.

Mei Xiang Ng, Thing Chai Tham, Sze Pheng Ong, Chung Lim Law *

University of Nottingham Malaysia Campus, Jalan Broga, 43500 Semenyih, Selangor, Malaysia

A R T I C L E I N F O

Article history:

Received 20 November 2014

Received in revised form

7 April 2015

Accepted 2 May 2015

Available online 11 May 2015

Keywords:

Crumb rubber

Drying characteristic

Mathematical modeling

A B S T R A C T

This paper reports the study of crumb rubber drying in different experimental designs. It is

important to understand the characteristics of crumb rubber drying in order to formulate a

better drying strategy that could give higher energy efficiency. Four experiments were car-

ried out with constant heat at maximum 100 �C and a stainless steel container was used to

hold the sample of crumb rubber under study. The surface temperature profile of the rub-

ber was investigated using two types of drying methods, normal hot air drying and vacuum

drying. It was found that when the sample was dried, external surface temperature for dry-

ing with hot air dryer was higher than vacuum dryer. The results showed the evolution of

temperature profile was not in good agreement with the prediction which revealed that

there was no temperature gradient within the drying samples. The energy consumption

for vacuum drying was higher compared to hot air drying, where there was a difference

of 0.7079 MJ/kg H2O evaporated for drying temperature at 100 �C. The best fit model gener-

ated from the experimental data was the modified Henderson and Pabis model and the

highest effective diffusivity obtained was 5.243 · 10�9 m2/s heating by vacuum oven at 90

�C under zero atmospheric pressure.

� 2015 China Agricultural University. Production and hosting by Elsevier B.V. All rights

reserved.

1. Introduction

Since early twentieth century, rubber has been an important

commodity for Malaysia [1]. The success in rubber planting

and the fast development of automobile industries have

made Malaysia one of the leading rubber exporters. Rubber

latex is the sap, a protective layer beneath the bark of rubber

tree (Hevea brasiliensis), which is normally tapped from the

bark of Hevea Brasiliensis tree, is also named as poly-

isoprene. This rubber tree originated from Brazil and its seed-

lings were then exported to Sri Lanka, Singapore and other

Asia countries, including Malaysia [2]. The early use of rubber

was restricted to waterproof shoes, it was then further popu-

larized when Charles Goodyear vulcanized the rubber into

modified rubber [3].

The current main rubber products for medical industries,

baby care and automotive industries are commonly produced

from concentrated latex or solid rubber. Both types of rubber

have to be further processed once the sap was tapped from

tree. The centrifugation of field latex will be able to produce

concentrated latex in liquid form with dry rubber content

(DRC) of 60% and above [4]; while rubber sheet or crumb rub-

ber is a type of dry rubber products that is rather important

for tyre industries. The raw rubber that was required for tyre

manufacturing process, known as crumb rubber, is com-

monly marketed in various grade and wrapped with polyethy-

lene plastic sheets [5]. Crumb rubber is also known as

‘‘technical specified rubber’’ [6], where 70% are sold to tyre

industries [7].

I n f o r m a t i o n P r o c e s s i n g i n A g r i c u l t u r e 2 ( 2 0 1 5 ) 6 4 –7 1 65

In this paper, the drying of technical specified rubber (TSR)

was investigated. In TSR, the initial plasticity, plasticity reten-

tion index and volatile matter are very important. For related

rubber products like gloves, the mechanical strength, includ-

ing tensile strength, elasticity, and modulus are part of the

crucial test to validate product quality. The rubber product’s

properties usually remain unchanged at normal room tem-

perature, but the strength, elasticity and flexibility of rubber

deteriorate faster when stored in hot condition. Generally,

crumb rubber was graded according to Standard Malaysia

Rubber (SMR), where the requirement was as mentioned in

Table 1.

Crumb rubber drying process is important in minimizing

the moisture content in rubber, reduction of water activity,

and to ensure consistent end product quality. Researches by

Kulchanat [8] and Yodthong [9] have shown that the addition

of wood vinegar to the drying process able to reduce bacteria

growth and inhibits anti-fungal properties on rubber. Thus,

the main objective of this study is to understand the drying

kinetic of crumb rubber and changes of temperature profile

on rubber surface when drying under hot air. The current rub-

ber drying requires drying air temperature in the range of 130

�C and heating continuously up to three hours, but the pro-

cess carries a possibility of inconsistent product quality with

high energy consumption. The most popular artificial drying

is the use of trolley dryer by hot air convective drying method.

However, the skyrocketing fuel cost is one of the main chal-

lenges to current rubber processing plants. There are limited

researches on rubber drying process. Berthomieu [10],

Khongchana [11], Yutthana [12] and Suchonpanit [13] works

have shown some similarities in trying to simulate and design

industrial dryers. According to Khongchana [11], the specific

energy consumption is measured by the equation below:

SEC ¼ 2:6P

WMð Þ þP

Qhð ÞDtWw

ð1Þ

where

SEC = Specific energy consumption, MJ/kg water

evaporated;

Qh = Heat energy consumption, kW;

WM = Electrical energy consumption, kW;

WW = Weight of water evaporated out from the rubber, kg.

However, there are multiple differences in Jutarut’s [14]

equation to measure the specific energy consumption:

SEC ¼ 2:6P

Efanð Þ �P

Qhð ÞDtMw

ð2Þ

Where

Table 1 – Technical specification for Standard Malaysia Rubber (

Parameter

Dirt Retained on 44 Aperture (Max, % wt)Ash Content (Max, % wt)Volatile Matter (Max, % wt)Nitrogen Content (Max, % wt)Wallace rapid plasticity, PO (Min, %)Plasticity retention index, PRI (Min, %)

SEC = Specific energy consumption, MJ/kg water

evaporated;

Qh = Rate of Heat energy consumption, kW;

Efan = Rate of Electrical energy consumption of fan, kW;

MW = Weight of water evaporated out from the rubber, kg.

The calculation presented by Jutarut involved the rate of

electrical energy and heat energy consumption. From the

two equations mentioned, we can see obvious differences in

both, whereby one was calculated based on the addition of

heat energy and electrical energy, and other SEC was using

subtraction of heat energy for calculation. The differences

in calculation methods may lead to difficulties to compare

the SEC values from different researchers. Therefore, there

is a need to investigate the actual energy consumption in sim-

ilar drying system. Besides evaluation of drying kinetic and

specific energy consumption (SEC) of rubber drying process,

a total of eleven types of mathematical models were analyzed

to determine whether the drying technique was suitable for

industrial scale rubber processing.

2. Materials and methods

2.1. Raw materials and equipment

Fresh crumb rubber (Lien Rubber (M) Sdn. Bhd., Port Klang,

Malaysia) was acquired from a local natural rubber processing

company. The crumb rubber was coagulated and washed

prior to purchase; however, further removal of dirt was neces-

sary for higher accuracy of test. The rubber size was reduced

by creper shredder machine and the appearance was similar

to a long thread with diameter of 4 mm ± 1 mm. The rubber

would entangle and stick to each other upon heat treatment.

Therefore, proper selection of rubber, washed off surface

adhering dirt, and size are important for the accuracy of dry-

ing test.

The drying of rubber was carried out by two types of dry-

ers, universal lab oven (Memmert, UFB 500, Germany,

Evergreen Engineering & Resources) and vacuum oven (Tuff,

TVAC-92, Germany, Tech-Lab Scientific Sdn Bhd) with the

use of a stainless steel container 150 mm · 75 mm

· 75 mm made by Sphere Corporation. A schematic diagram

(see Fig. 1) showed how the rubber sample was placed into

the container. The universal oven employed hot air (HA) dry-

ing method by natural convection, with a heater of 1600watt

and temperature accuracy of ±0.1 �C; while the vacuum dryer

employed vacuum drying (VD), which consisted of a build in

SMR).

SMR 10 SMR 20

0.08 0.160.75 1.000.80 0.800.60 0.6030 3050 40

Fig. 1 – Schematic diagram of rubber in container.

66 I n f o r m a t i o n P r o c e s s i n g i n A g r i c u l t u r e 2 ( 2 0 1 5 ) 6 4 –7 1

vacuum pump, temperature controlling unit, and a heater of

2400 watt with accuracy ±0.1%. An analytical weighing bal-

ance (Mettler Toledo, ME204, Malaysia, Mettler-Toledo (M)

Sdn Bhd) with accuracy up to ±0.0001 g was utilized to mea-

sure the weight changes of rubber compound throughout

the experiment, while the drying time for all experiments

were determined by using a digital sport timer (Avantec,

TM-104, Malaysia, Suria Pembekal Umum Sdn Bhd).

2.2. Process control and design

The crumb rubber sample was thoroughly cleaned and

soaked in cold water for ten minutes prior to drying to obtain

a saturated rubber sample. Then, the rubber was cut into

small pieces to allow uniform heat distribution on the sam-

ples, and to fit into the stainless steel container. The rubber

sample subjected to drying had a dimension of (average

length, width and height) 150 mm · 75 mm · 30 mm. The

mass of the rubber was determined before drying and

throughout the heating process with increasing interval per-

iod. The rubber samples were heated at a pre-set drying tem-

perature in the range of 80–100 �C and the process was

continuously monitored up to 65 min. Uniformity in drying

was maintained throughout the study to determine the effect

of heating temperature on rubber, including weight change

and moisture content. The dried samples were then cooled

to room temperature after each drying experiment, and

sealed in polyethylene bags for future references. The influ-

ence of dryer on temperature profile and specific energy con-

sumption was determined from the experiment carried out.

Some measurements were taken prior to drying such as ini-

tial moisture content and mass of each sample at 65% (dry

basis) and �30 ± 5 g, respectively. The moisture content of

fresh sample and dried sample were obtained by the AOAC

method No. 934.06. The average ambient room temperature

and relative humidity were recorded at 27 �C and 80%,

respectively.

2.3. Moisture content, moisture ratio and drying rate

In order to understand the fundamental drying process in

rubber, the measurement of pre-set drying temperature,

ambient room temperature, rubber’s surface temperature,

and initial moisture content were made. The average initial

moisture content in rubber was 65 ± 1%, while wet basis

moisture was 39 ± 1% and the data was continuously recorded

for 65 min. There were four drying conditions being evaluated

as shown in Table 2. The dry basis moisture content, MC was

calculated by using the below equation:

Moisture Content ðgrams water=gram dry solidÞ

¼ Weight of waterWeight of dry solids present

ð3Þ

During the drying process, all temperature data were mea-

sured with an infrared thermometer, with an accuracy of ±2

�C. The drying rate was determined from the changes of mass

with time, while drying constant, k was typically obtained

from the slope of the negative natural log of the moisture

ratio, ln (MR) versus time. The moisture ratio, MR was calcu-

lated from the results obtained by following Fick’s second law

[16] in the theoretical model of thin layer drying:

MR ¼ M�Me

Mi �Með4Þ

Where MR is moisture ratio, Mi is initial moisture content,

Me is equilibrium moisture content, and M is moisture con-

tent at time t.

The drying kinetic and drying rate for rubber drying were

calculated based on the data obtained. The drying rate, W (g

H2O/m2 min) was defined by:

W ¼ �Wd

AdXdt

� �ð5Þ

where W is drying rate, Wd is weight of dry solid (Bone dry

mass), A is the contact surface of the drying gas and the rub-

ber, and dX/dt is the humidity variation (dry basis) over time t.

Table 2 – Experimental design for drying of rubber.

Code Heat source Drying temp. (�C) Drying time (min) Atmospheric pressure (atm)

T1 Vacuum oven 80 65 0T2 Vacuum oven 90 65 0T3 Vacuum oven 100 65 1T4 Universal oven 100 65 1

**T4 was conducted to imitate current industrial drying by hot air convective method [15].

I n f o r m a t i o n P r o c e s s i n g i n A g r i c u l t u r e 2 ( 2 0 1 5 ) 6 4 –7 1 67

2.4. Specific energy consumption (SEC)

SEC denoted the energy required to remove a unit mass of

water. For the calculation of SEC, the equation used as

reference is the equation from the study of Bualuang [17].

The calculation based on this equation could adequately

compare the SEC of two different dryers, while the two equa-

tions of Khongchana [11] and Jutarut [14] mentioned earlier

can only calculate the SEC based on water evaporated, heat

energy and electrical energy consumption without consider-

ing the bone dry mass of sample.

SEC ¼ 3:6PðMi �MfÞWd

ð6Þ

where P is power, Mi is initial moisture content, Mf is final

moisture content; and Wd is bone dry mass of sample, kg.

2.5. Drying kinetic models

By evaluating the drying kinetic of rubber, the efficiency of

vacuum drying and hot air convective drying method was

analyzed. The analysis allowed the determination of suitable

drying technique that was suitable for industrial scale rubber

processing. There were eleven types of mathematical models

to describe the drying kinetics of rubber (see Table 3). Some of

the earlier models were based on thin layer rapid drying with

low temperature heating profile, which was not suitable for

applying on thick layer drying experimental works [18]. The

complex partial differential equation that involved mass

balance, drying rate, heat balance and transfer would be

more accurate, but a model must be validated by comparison

with experimental results. The model that was able to

describe rubber drying characteristics would give the best fit

Table 3 – Mathematical modeling applied to the drying curves.

Model No. Model name Model equatio

1 Lewis MR = exp(�kt)2 Page MR = exp(�ktn)3 Logistic MR = a/(1 + exp4 Two term MR = a exp(�k5 Wang and Singh MR = 1 + at + b6 Henderson and Pabis MR = a exp(�k7 Logarithmic MR = a exp(�k8 Approximation of diffusion MR = a exp(�k9 Two term exponential MR = a exp(�k

10 Modified Henderson and Pabis MR = a exp(�k11 Hii model MR = a exp(�k

Note: k, a, b, c, g and n are drying constants.

on the experimental data. The correlation coefficient (R2),

chi-square and RMSE were used to determine the quality of

the fit. The highest value of R2, where the value greater than

0.995 indicating a good fit. The model with lowest values of

chi-square and RMSE were chosen as the best empirical

model equation for rubber drying.

3. Results and discussions

3.1. Drying curve

From the rubber drying studies, the temperature profile

through changes of rubber’s surface temperature (see Fig. 2)

was investigated. For initial heating process, the heat transfer

was fast as the moisture content of rubber was high. As the

evaporation process of water was exothermic, heat was

absorbed quickly by the rubber when both tests were initiated

under setting temperature 100 �C. For drying under vacuum

condition, the slow rising of rubber’s surface temperature

indicated that the heat conduction to the rubber sample

was not effective. As rubber was not a good heat conductor,

the conduction of heat through contact surface would be

slow. The highest rubber surface temperature measured

throughout the test was only 74 �C, which indicated that there

was heat loss during thermal diffusion in the rubber.

However, a total of 10.998 g moisture was able to be removed

due to the intense heat provided by the vacuum oven

compared to universal oven. For the drying test carried out

by universal oven, the dryer provided sufficient air flow to

spread the heat evenly on rubber surface, with higher impact

on rubber’s surface temperature compared to vacuum drying,

but the total moisture loss was 8.091 g only. The drying in vac-

uum condition was faster compared to hot air drying by

n References

Bruce (1985)Page (1949)

(�kt)) Apintanapong (2009)

1t) + b exp(�k2t) Henderson (1974)t2 Wang and Singh (1978)t) Henderson and Pabis (1961)t) + c Yaldiz et al. (2001)t) + (1 � a) exp(�kbt) Yaldiz et al. (2001)t) + (1 � a) exp(�kat) Akpinar et al. (2003)

1t) + b exp(�k2t) + c exp(�k3t) Tasara et al. (2011)tn) + c exp(�gtn) Hii et al. (2009)

Table 4 – Drying rate and specific energy consumption (SEC) analysis.

Exp Heat Source Drying temp. (�C) Dryingtime (min)

Averagedrying rate(g H2O/m2 min)

Deff (m2/s) P (kW-h) SEC (MJ/kg H2Oevaporated)

1 Vacuum oven 80 65 15.8437 3.716 · 10�9 9.360 9.10952 Vacuum oven 90 65 13.4385 5.243 · 10�9 9.360 9.70953 Vacuum oven 100 65 16.5931 4.537 · 10�9 9.360 8.51034 Universal oven 100 65 10.3529 3.830 · 10�9 6.239 7.8024

Fig. 2 – Temperature profile with respect to time.

68 I n f o r m a t i o n P r o c e s s i n g i n A g r i c u l t u r e 2 ( 2 0 1 5 ) 6 4 –7 1

referring to the moisture removal rate and calculated drying

rate (see Table 4).

For moisture removal according to time interval, the water

removal rate was rapid in the first ten minutes and deceler-

ated as moisture content started to reduce. As the surface

moisture dried out, the moisture need longer time to diffuse

out from rubber. The stickiness of the rubber inhibited the

diffusion of internal moisture to rubber surface, but the heat

would constantly soften the rubber, making it more compact

as compared to the time when it was moist. The changing of

moisture ratio (MR) versus time (see Fig. 3) showed that MR

decreased continuously throughout 65 min and no constant

drying rate period existed. It is apparent that the drying

became slower after the first drying period and falling-rate

period was present. As the moisture was removed from rub-

ber, the rubber got thinner and experienced slight shrinkage

throughout the drying period, it was predicted that the drying

Fig. 3 – Moisture ratio with respect to time.

condition (including humidity, temperature and air flow)

throughout the rubber were constant.

Drying kinetics was studied for moisture contents 65–20%

(w/w). Four drying curves of rubber with drying temperature

of 80–100 �C versus moisture content were plotted (see

Fig. 4). All drying condition showed a trend of reducing drying

rate, this was due to the increase of rubber’s bulk density and

the reduction of pore for air flow. The initial moisture content

(MC) of all samples were maintained at 65% and all tests were

able to get results lower than 20% dry basis MC within 65

min. This result showed that the rubber would be able to

dry to desired moisture content at shorter given time, instead

of three hours as per industrial practice. From Fig. 4, it can be

seen that falling-rate period began after the warm up time

and subsequent drying process by universal oven was slower

as compared to vacuum oven. The experimental work done

by using universal oven also obtained results with higher

MCf compared to vacuum oven.

3.2. Specific energy consumption (SEC)

The results demonstrated that the drying rate by vacuum

oven at 100 �C was highest among all the four tests (see

Table 4) even though the drying process was operated at 1

atmospheric pressure. This showed that rubber was more

suitable to dry in a condition with hot air as medium com-

pared to in vacuum state. However, the SEC results showed

that universal oven had lowest MJ/kg H2O evaporated, the

energy consumption for vacuum dryer was higher compared

to hot air dryer, which agreed to same findings on other

research work for different materials [19].

Fig. 4 – Drying rate (DR) versus % moisture content (MC) (a) vacuum oven 80 �C; (b) vacuum oven 90 �C; (c) vacuum oven 100 �C;

(d) universal oven 100 �C.

I n f o r m a t i o n P r o c e s s i n g i n A g r i c u l t u r e 2 ( 2 0 1 5 ) 6 4 –7 1 69

3.3. Drying kinetic models

From the comparison of eleven mathematical models

mentioned, modified Henderson and Pabis Model was the

most suitable empirical model (see Table 5), as it had high

R2 value of minimum 0.9979 and lowest chi-square value of

maximum 0.002. The arbitrary constant for the drying

condition was as described (see Table 6). This empirical

modified Henderson and Pabis model has shown a better fit

to the experimental rubber drying data as compared to others

models. Consequently, the modified Henderson and Pabis

drying model could adequately describe the drying

behavior of rubber, a good correlation between the values of

Table 5 – Modeling of moisture ratio versus drying time of rubb

Model name T1 T2

R2 Chi-sq RMSE R2 Chi-s

1. Lewis 0.3617 2.7311 0.1902 0.7034 0.7782. Page 0.9662 0.029 0.3108 0.9953 0.0063. Logistic 0 1.0082 0 0.851 0.2154. Two term 0.9975 0.0027 0.3158 0.9958 0.0075. Wang and Singh 0 2.9576 0 0.6939 0.3596. Henderson and Pabis 0.6902 0.3437 0.2627 0.8886 0.1387. Logarithmic 0.9629 0.0498 0.3103 0.9389 0.0588. Approximationof diffusion

0.9974 0.0027 0.3158 0.9958 0.007

9. Two term exponential 0.5441 1.2614 0.2333 0.8103 0.37510. ModifiedHenderson and Pabis

0.9979 0.002 0.3159 0.9996 0.000

11. Hii model 0.9976 0.0028 0.3159 0.9993 0.001

Lowest 0 0.002 0 0.6939 0.000Highest 0.9979 2.9576 0.3159 0.9996 0.778

model and experiment data was found. The predicted

moisture ratio versus experimental moisture ratio graph for

four drying condition was presented in four charts (see

Fig. 5). From the slope of graph ln (MR) versus time, the value

of effective moisture diffusivity was calculated and the

value of Deff for all four drying condition were presented

in Table 4.

4. Conclusions

The drying curve revealed that rubber drying process involved

multiple stages. After the initial warm up period, the drying

process continued in falling rate period and no constant rate

er.

T3 T4

q RMSE R2 Chi-sq RMSE R2 Chi-sq RMSE

5 0.2652 0.6224 0.7121 0.2495 0.5678 1.1252 0.23837 0.3155 0.9825 0.0144 0.3135 0.9988 0.0017 0.3165 0.2917 0.7586 0.2523 0.2754 0.7871 0.2457 0.28066 0.3156 0.9994 0.0007 0.3161 0.9916 0.0092 0.31494 0.2634 0.7206 0.2735 0.2684 0.5976 0.4259 0.24451 0.2981 0.8117 0.1998 0.2849 0.8324 0.1823 0.28859 0.3064 0.9733 0.0303 0.312 0.9309 0.0674 0.30516 0.3156 0.9994 0.0007 0.3161 0.9915 0.0096 0.3149

9 0.2847 0.7594 0.3891 0.2756 0.7089 0.5674 0.26634 0.3162 0.9997 0.0003 0.3162 0.9997 0.0004 0.3162

1 0.3161 0.9994 0.0006 0.3161 0.9996 0.0004 0.3162

4 0.2634 0.6224 0.0003 0.2495 0.5678 0.000368 0.23835 0.3162 0.9997 0.7121 0.3162 0.9997 1.1252 0.3162

Fig. 5 – Comparison of moisture ratio between experimental data and calculated values according to modified Henderson and

Pabis model (a) vacuum oven 80 �C; (b) vacuum oven 90 �C; (c) vacuum oven 100 �C; (d) universal oven 100 �C.

Table 6 – Arbitrary constant of four drying condition by modified Henderson and Pabis Model.

Arbitrary constant/drying condition Henderson and Pabis Model

T1–VA 80 �C T2–VA 90 �C T3–VA 100 �C T4–HA 100 �C

k1 0.005690974 0.014635181 0.014090465 0.006861388a 0.376967754 0.479935538 0.542363668 0.44701359b 0.503064087 0.266047652 0.19188322 0.40640397k2 0.491953271 1.641019116 5.80132881 0.322075361c 0.127525422 0.25393464 0.265755397 0.148017407k3 0.043057648 0.076591971 0.224971683 0.047721102

70 I n f o r m a t i o n P r o c e s s i n g i n A g r i c u l t u r e 2 ( 2 0 1 5 ) 6 4 –7 1

period existed. The falling rate was due to the shrinkage of

material and increase in bulk density. The study also revealed

that the drying rate changes with different dryers’ perfor-

mances, whereby hot air drying method consumed less heat

energy with less moisture removal in the process. The specific

energy consumption of hot air drying showed that it can

reduce the energy consumption for rubber drying. However,

more heating time was required when hot air drying method

was employed in the drying process.

The temperature profile of rubber showed the effective-

ness of heat transfer from surrounding to rubber. The uni-

versal oven was able to distribute heat faster compared to

vacuum oven. The drying kinetics was mainly affected by

drying temperature, where higher temperature improved

the drying rate and reduction of moisture content in rubber

sample. The drying rate was higher in vacuum condition

due to the condition of lower atmospheric pressure, so dry-

ing in vacuum condition was able to reduce MC in higher

rate. The empirical model named modified Henderson and

Pabis model were the best fit model for the measured

experimental values.

Acknowledgements

The authors wished to acknowledge the contributions made

by Sphere Corporation Sdn Bhd, University Of Nottingham –

Malaysia Campus (UNMC) for their financial support.

R E F E R E N C E S

[1] Drabble JH. An economic history of Malaysia, c.1800–1990: thetransition to modern economic growth. Economic HistoryAssociation; 2001. p. 1–2.

[2] Kinnaman B. A brief natural history of latex rubber allergy.Where does rubber come from? 1997 [cited 2014 6th April].

[3] Steven T, Carlos M, Zephyr F. From silver to cocaine: latinamerican commodity chains and the building of the worldeconomy, 1500–2000. In: Zephyr Frank AM, editor. Brazil inthe international rubber trade. Duke University Press; 2006.

[4] Santipanusopon S, Riyajan S-A. Effect of field natural rubberlatex with different ammonia contents and storage period onphysical properties of latex concentrate, stability of skimlatex and dipped film. Phys Procedia 2009;2:127–34.

I n f o r m a t i o n P r o c e s s i n g i n A g r i c u l t u r e 2 ( 2 0 1 5 ) 6 4 –7 1 71

[5] Utomo TP, Hasanudin U, Suroso E. Comparitive study of lowand high-grade crumb rubber processing energy. In:Proceedings of the world congress on engineering 2010, volIII; 2010.

[6] Setyamidjaja D. Karet Seri Budi Daya. Yogyakarta: Kanisus;1993.

[7] Cousin B, Benet JC, Auria R. Experimental study of the dryingof a thick layer of natural crumb rubber. Drying Technol1993;11(6):1401–13.

[8] Kulchanat P, Namthip R, Juraivan R. Effects of wood vinegaras an additive for natural rubber products. Songklanakarin JSci Technol 2011;33(4):425–30.

[9] Yodthong B, Noi N. Study on wood vinegars for use ascoagulating and antifungal agents on the production ofnatural rubber sheets. Biomass Bioenergy 2009;33:994–8.

[10] Berthomieu G, Sle Beuve J, Lelong T, Benet JC, Saix C. Fromsimulation to the design of an industrial drier for naturalrubber in granule form. Drying Technol Int J1997;15(10):2541–55.

[11] Khongchana P, Supawan T, Yutthana T, WoravutthikhunchaiS. Effect of drying strategies on quality of STR 20 blockrubbers. J Agr Technol 2007;3(2):157–71.

[12] Yutthana T, Supawan T. Mathematical model of fixed-beddrying and strategies for crumb rubber producing STR20.Drying Technol Int J 2008;26(11):1388–95.

[13] Suchonpanit W, Tirawanichakul Y, Tirawanichakul S.Experimental study of the effects of different drying methodson the kinetics and quality of air dried sheet (ADS) rubber. In:CHUSER 2011. Penang; 2011.

[14] Jutarut T, Yutthana T, Supawan T. Effect of dryingcondition on natural block rubber: simulation andExperiments. In: 2nd Thammasat University internationalconference on chemical, environmental and energyengineering; 2009.

[15] Intapun J. The effects of biological maturation of coagula ofHevea Brasiliensis latex on dry rubber properties [Ph.D.thesis]. Prince of Songkla University: Thailand; 2009.

[16] Crank J. The mathematics of diffusion. London: ClarendonPress; 1975.

[17] Bualuang O, Tirawanichakul Y, Trawanichakul S. Dryingkinetics study of parboiled rice by using artificial neuralnetwork model. In: CHUSER 2011. Penang; 2011.

[18] Sharp JR. A review of low temperature drying simulationmodels. J Agr Eng Res 1982;27:169–90.

[19] Nadee A, Tirawanichakul Y, Tirawanichakul S. Drying kineticmodel and physical properties of herb pandanus leaf. In: IEEEcolloquium on humanities, science and engineering research(CHUSER). Penang; 2011. p. 54–59.